Microsoft Principal Applied Scientist 

Taiwan, Taoyuan City 

374309328

As a

Required Qualifications:

• Bachelor's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 12+ years related experience (e.g., statistics, predictive analytics, research)
  • OR Master's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 10+ years related experience (e.g., statistics, predictive analytics, research)
  • OR Doctorate in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 8+ years related experience (e.g., statistics, predictive analytics, research)
  • OR equivalent experience.
• At least 3+ years of experience delivering team level outcomes.
  
• 2+ years of industrial experience coding in C++, C#, C, Java or Python.
  
• Prior experience with data analysis or understanding, looking at data from a large-scale systems to identify patterns or create evaluation datasets.
  
• Familiarity with common machine learning, deep learning frameworks and concepts, using use of LLMs, prompting.
  
• Ability to communicate technical details clearly across organizational boundaries.

Preferred Qualifications:

• 5+ years experience creating publications (e.g., patents, libraries, peer-reviewed academic papers).
• 2+ years experience presenting at conferences or other events in the outside research/industry community as an invited speaker.

• Driving projects from design through conception, implementation, experimentation and finally shipping to our users. This requires deep diving into data to identify gaps, coming up with heuristics and possible solutions, using LLMs to create the right model or evaluation prompts, and setting up the engineering pipeline or infrastructure to run them.
• Documenting progress & processing, assisting & guiding junior team members, aligning & unblocking them with other stakeholders in timezones.
• Coming up with evaluation techniques, datasets, criteria and metrics for model evaluations. These are often SOTA models or metrics / datasets.
• Hands on pre-training, fine-tuning, use of language models, including dataset creation, filtering, review, and continuous iteration. This may also require understanding of training frameworks, formats, checkpoints, stacks such as megatron.